Pith. sign in

REVIEW 3 major objections 4 minor 55 references

Estimation with missing not at random binary outcomes via exponential tilts

T0 review · 3 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Exponential tilt models identify missing-not-at-random binary outcomes from the observed marginal.

desk verdict A genuinely new identifiability condition and a clean DR theorem, but the Waterbirds real-data claim rests on an unverified assumption that the paper waves through. read the letter →

arxiv 2502.06046 v1 pith:36HWDM5O submitted 2025-02-09 stat.ME

Subha Maity This is my paper · ORCID
classification stat.ME MSC 62D1062F12
keywords missingnotatrandombinaryoutcomesexponentialtiltimportanceweightingdoublyrobustestimationidentifiabilitymissingnesscontrollingvariableunsupervisedtransferlearning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that, for binary outcomes missing not at random (MNAR), the missingness problem is identifiable when the importance weight between the responder and non-responder subpopulations has an exponential tilt form, $\omega(x,y)=\exp(\alpha_y+\beta_y^\top T(x))$. Under one sufficient condition on the observed-population log odds (Assumption 2.1), the tilt parameters are uniquely determined by the observed marginal distribution of $X$ alone, so the method needs neither nonresponse instruments nor shadow variables. The paper estimates the tilt parameters by a constrained Kullback-Leibler matching problem and feeds them into importance-weighted and doubly robust estimators for any mean functional. If the central claim is right, routine observed data with binary outcomes can support transfer learning and missing-data adjustment in settings where previous MNAR methods required external identifying variables.

What carries the argument

The load-bearing object is the exponential tilt model (1.2), a two-component exponential-family shift $\omega(x,y)=\exp(\alpha_y+\beta_y^\top T(x))$ between the $R=1$ and $R=0$ subpopulations. This single specification carries the whole argument: it turns the observed covariate distribution into a two-component tilted mixture (eq. (2.2)), which is matched to the data by minimizing Kullback-Leibler divergence under a normalization constraint (eqs. (3.1)-(3.2)); it delivers the propensity score formula in Lemma 3.1; and its log-odds difference $\log\omega(x,1)-\log\omega(x,0)=(\alpha_1-\alpha_0)+(\beta_1-\beta_0)^\top T(x)$ provides the outcome regression in Lemma 3.2. The double-robustness theorem then rests on those two building blocks.

What would settle it

Construct data in which the $R=1$ log odds has exactly the forbidden form, for example by taking two Gaussian components whose means are swapped as in the paper's Section 2 counterexample. Then run Algorithm 1 from multiple starting points: the objective (3.2) has multiple optima, so the recovered importance weights will differ across runs, directly contradicting the claim of unique identification.

Watch

Extended reading notes

Core claim

Under the exponential tilt model (1.2), the joint law of $(X,Y)$ in the missing domain is obtained from the observed domain by multiplying each outcome-specific density by $\exp(\alpha_y+\beta_y^\top T(x))$; therefore the observed covariate marginal satisfies eq. (2.2). The paper shows that Assumption 2.1 selects a unique set of tilt parameters from that marginal, making the importance weights, and hence all mean functionals, identifiable. The mechanism is complete: the missingness propensity score is $\mathrm{P}(R=1\mid X,Y)=\mathrm{P}(R=1)/(\mathrm{P}(R=1)+\omega(X,Y)\mathrm{P}(R=0))$, the conditional outcome probability under missingness is obtained from a log odds ratio $(\alpha_1-\alpha_0)+(\beta_1-\beta_0)^\top T(x)$, and the doubly robust estimators are asymptotically unbiased if either the tilt model is correct or the outcome regression plus log-odds-ratio model is correct (Theorem 3.4). The paper additionally reports that the fitted linear combination $(\beta_0^\top T(X),\beta_1^\top T(X))$ acts as an estimated missingness controlling variable, and that in the Waterbirds study this suffices for unsupervised transfer learning at accuracy comparable to a labeled target oracle.

Load-bearing premise

The entire identification rests on Assumption 2.1: the true log odds $\log\{\mathrm{P}(Y=1\mid X=x,R=1)/\mathrm{P}(Y=0\mid X=x,R=1)\}$ must not be representable by the forbidden log-ratio-of-exponentials expression, otherwise two different tilt parameter values explain the observed covariate marginal equally well.

Editorial extensions

If this is right

  • Mean functionals $\mathrm{E}[\tau(X,Y)]$ and $\mathrm{E}[\tau(X,Y)\mid R=0]$ become estimable from observed data alone whenever the tilt parameters are identified.
  • A user no longer needs to pre-specify a nonresponse instrument or shadow variable; the procedure returns an estimated missingness controlling variable $(\beta_0^\top T(X),\beta_1^\top T(X))$ as a byproduct.
  • Unsupervised transfer learning when the outcome-given-covariate relationship differs between source and target can be carried out without labeled target samples, as long as the source-population log odds satisfy Assumption 2.1.
  • The doubly robust estimator keeps asymptotic unbiasedness under misspecification of the tilt model provided the log-odds-ratio and outcome regression are correct, a robustness property shown in Theorem 3.4.
  • A computationally stable exponentiated-gradient algorithm solves the constrained KL matching problem, outperforming a profile empirical likelihood alternative in the paper's simulations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper asserts Assumption 2.1 is in principle verifiable but gives no practical test; fitting the observed log odds nonparametrically and checking whether it ever enters the forbidden log-ratio family would be a natural way to make the method safer in practice.
  • Because the estimated missingness controlling variable is only identified up to the choice of summary $T(x)$, the surrogate may recover a transformation of the true controlling variable rather than the variable itself; this is enough for prediction but not automatically for causal interpretation.
  • The Waterbirds experiment suggests the DR estimator can become unstable when the product of estimated residuals and covariate directions is unbounded, so a clipped or regularized variant of the DR loss is a natural extension.
  • The same tilt machinery should extend to multi-category outcomes or to $T(x)$ chosen as a neural embedding, as in the Waterbirds experiment, but the identifiability condition would need to be re-derived for each new family.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. This paper studies estimation of mean functionals when binary outcomes are missing not at random. The central assumption is that the importance weight mapping the observed (R=1) joint distribution to the missing (R=0) joint distribution has the exponential-tilt form exp(alpha_y + beta_y' T(x)). Under a new sufficient identifiability condition (Assumption 2.1), the tilt parameters are claimed identifiable from the marginal distribution of X conditional on R=0 and R=1. The paper proposes a constrained KL-matching estimator solved by an exponentiated-gradient algorithm, constructs importance-weighted and doubly robust estimators for mean functionals, and proves a double-robustness property. Synthetic normal-mixture experiments support the theory, and a Waterbirds transfer-learning experiment is used to argue that the method mitigates spurious correlation without observing the missingness-controlling variable.

Significance. If the claims hold, the paper offers a genuinely useful direction: it weakens the usual requirements of nonresponse instruments or shadow variables for MNAR binary outcomes, and it provides a double-robustness result that can tolerate certain misspecifications of the tilt model. The internal theoretical core is largely consistent: Lemma C.4, Lemma 3.3, and Theorem 3.4 are coherent, the bias decompositions are explicit, and the synthetic experiments match the claimed behavior. The paper also ships code and evaluates the transfer-learning claim on a held-out target-test split with oracle benchmarks. However, the practical usefulness of the method rests on Assumption 2.1, and the only real-data demonstration does not verify that assumption. In addition, the estimation algorithm lacks convergence guarantees, so the statistical claims are not fully supported.

major comments (3)
  1. [Section 5, Assumption 2.1] The sentence 'The functional structure of single hidden layer neural network satisfies the Assumption 2.1' does not establish the identifiability condition. The forbidden family in Assumption 2.1 contains every affine function, since taking mu0=mu1 and delta0=delta1=0 gives log odds equal to mu2 + delta2' t. A single-hidden-layer neural network is a universal approximator on compact covariate support, so the network class is not structurally disjoint from the forbidden family. What would need to be checked is whether the fitted log-odds logit(eta1(x)) learned from the Waterbirds source data is not close to the forbidden family on the support of X. No such check is reported. Because identification of theta through equation (2.2) is entirely conditional on Assumption 2.1, the reported IW and OR accuracy gains in Figure 3 are not established as identifiable quantities; they could reflect an arbitrary selection among near-equivalent theta values.
  2. [Section 3.1, Algorithm 1] No convergence, stationarity, or local-minimum guarantee is given for the exponentiated-gradient algorithm. The objective in (3.2) is nonconvex in theta because of the exponential terms and the estimated nuisance function eta1, and the Lagrangian update in Algorithm 1 is heuristic. Lemma 3.3 and Theorem 3.4 are asymptotic bias statements that assume thetahat -> theta* and xihat -> xi*, but no theorem establishes these limits for the estimator produced by Algorithm 1. Please either provide conditions under which Algorithm 1 converges to a consistent estimator, or explicitly state the consistency claim as an assumption/conjecture and adjust the wording of Theorem 3.4 accordingly.
  3. [Section 2, Assumption 2.1] The paper states that Assumption 2.1 is 'in principle verifiable from the observed dataset', but it offers no practical verification procedure. The quadratic-log-odds example is only a sufficient instance, not a test of non-membership in the forbidden family. Since this assumption is the sole route to identification of the tilt parameters, the manuscript should either provide a falsifiable diagnostic for the assumption (for example, a goodness-of-fit test against the forbidden family on the R=1 sample) or clearly frame the real-data analysis as illustrative under an unverified condition. Without this, the identification claim in the Waterbirds study is not supported.
minor comments (4)
  1. [Section 5, Eq. (5.1)] In the definition of the outcome-regression loss, the second term should be (1-p) times the loss at Y=0; as written, ell(X,p;beta) = p ell(X,1;beta) + (1-p) ell(X,1;beta) is not the expected loss under a probability p.
  2. [Algorithm 1, line 9] The indicator condition for updating eta2 has a missing closing parenthesis: 'I{gn(theta(t) < -epsilon}' should read 'I{gn(theta(t)) < -epsilon}'.
  3. [Section 4, paragraph after Eq. (4.1)] The text identifies the target as mu0 = pi_{1|r} = 0.6; this should be pi_{1|0} = 0.6, since mu0 = E[Y | R=0].
  4. [Appendix C.2.4, proof of Theorem 3.4] The proof contains a duplicated, incomplete sentence: 'leading to bias(hat(mu)) = 0. This establishes (i). To establish (ii)' appears before the decomposition in (C.8), and should be removed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the tilt parameters are estimated from observed R=0 marginals and R=1 labels, and the target-test evaluation is held out; the main theorems are proved in-appendix from stated assumptions.

full rationale

Walking the derivation chain: identifiability (Assumption 2.1 plus Lemma C.4) is a sufficient condition proved from the model equations, not imported from prior work. Estimation in (3.2) minimizes KL(p(x|R=0), sum_y exp(alpha_y+beta_y^T t)p(x,y|R=1)) using only observed X in R=0 and the fitted source classifier eta_1; no missing Y values from R=0 enter the fitting objective. The IW and DR estimators are plug-in constructions, and Theorem 3.4's bias expressions are derived in Lemmas 3.3 and C.7 rather than assumed. In the Waterbirds experiment, target-train is used only to estimate the tilt parameters and target-test is used only for evaluation, with oracle benchmarks; no fitted quantity is presented as a prediction of itself. The one potentially circular-looking sentence, 'The functional structure of single hidden layer neural network satisfies the Assumption 2.1,' is not a construction-level circularity but an unsupported identifiability claim: Assumption 2.1 is a non-membership condition on the true source log-odds, and the neural-network functional class is not structurally disjoint from the forbidden family (constant functions are in the family). That is a correctness or verification risk, not a reduction of the result to its inputs. Self-citations to Maity et al. (2023, 2024) supply terminology and background; they do not carry the derivation. The paper is self-contained with proofs in the appendix, so no circular step is exhibited.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central claim rests on the exponential tilt model, the sufficient identifiability condition, correct specification of the nuisance model η1, and the heuristic convergence of Algorithm 1. No new physical entities are postulated; the estimated MCV surrogate U is a derived projection of X. The most fragile input is Assumption 2.1, which is hard to verify in finite samples and is merely asserted for Waterbirds.

free parameters (4)
  • tilt parameters θ = (α0, α1, β0, β1) = estimated by Algorithm 1; β in R^2 (synthetic) and R^512 (Waterbirds)
    The core estimated quantities. All importance weights, IPW, DR estimators, and the MCV surrogate depend on them.
  • nuisance classifier parameters for η1 (random forest / single hidden layer NN) = fitted on R=1 data; values not reported
    Enter the KL objective (3.2) and the outcome regression (3.10); misspecification biases IW and breaks DR condition (ii).
  • sufficient statistic T(x) = T(x)=x in both experiments
    A modeling choice fixing the tilt family; the identifiability condition and the linear log-odds assumption depend on it.
  • Algorithm 1 hyperparameters (B, ρ1, ρ2, ε, λ, tol, max_iter) = B=5, eps=1e-3, tol=2e-3, lr=4e-3, max_iter=4000, reg=1e-5 (from code signature)
    Chosen by hand in the implementation; influence the estimated θ and downstream results; no sensitivity analysis reported.
assumptions (6)
  • domain assumption Exponential tilt model (2.1): p(x,y|R=0) = exp(α_y + β_y^T t) p(x,y|R=1)
    Central parametric assumption stated in eq. (1.2) and (2.1). If violated, IW estimators are generally biased; DR has partial robustness only for the γ functional.
  • ad hoc to paper Assumption 2.1 (sufficient identifiability condition on the log odds)
    Ensures a unique solution for θ from the observed marginal in eq. (2.2). The paper calls it verifiable in principle but provides no practical test and does not verify it for Waterbirds.
  • standard math Existence of densities for P(X|Y=y,R=r)
    Invoked in Section 2 for the density ratio formulation of the tilt model and throughout the proofs.
  • domain assumption Correct specification of the nuisance model η1 for IW consistency and DR condition (ii)
    The KL objective (3.2) and outcome regression (3.10) use an estimate of P(Y=1|X,R=1); Lemma 3.3 and Theorem 3.4 assume consistency or well-specification of this model.
  • domain assumption Convergence of Algorithm 1 to the global minimizer of (3.2)
    The exponentiated gradient algorithm is heuristic; no convergence guarantee is stated, and the objective is non-convex with a nonlinear constraint.
  • domain assumption Subpopulation shift in Waterbirds: X ⊥ R | Y, A with A the background, and linear log-odds approximation by T(x)=x
    Section 5 describes the data generating process. The claim that the tilt model approximates the MCV is validated empirically, not theoretically.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Estimation with missing not at random binary outcomes via exponential tilts." pith.science (2026). https://pith.science/paper/36HWDM5O

@misc{pith2026250206046,
  author       = {Pith},
  title        = {Pith review of: Estimation with missing not at random binary outcomes via exponential tilts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/36HWDM5O}},
  note         = {Machine review of arXiv:2502.06046}
}
read the original abstract

We study the problem of missing not at random (MNAR) datasets with binary outcomes. We propose an exponential tilt based approach that bypasses any knowledge on 'nonresponse instruments' or 'shadow variables' that are usually required for statistical estimation. We establish a sufficient condition for identifiability of tilt parameters and propose an algorithm to estimate them. Based on these tilt parameter estimates, we propose importance weighted and doubly robust estimators for any mean functions of interest, and validate their performances in a synthetic dataset. In an experiment with the Waterbirds dataset, we utilize our tilt framework to perform unsupervised transfer learning, when the responses are missing from a target domain of interest, and achieve a prediction performance that is comparable to a gold standard.

Figures

Figures reproduced from arXiv: 2502.06046 by the authors.

Figure 1
Figure 1. Different estimates in well-specified exponential tilt model in [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Different estimates in misspecified exponential tilt model in [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Waterbirds case-study. The left plot presents proportions of different groups in source and target domain and the right plot provides performances of different prediction models on the target domain. where ℓ(X,Y; β) is the loss for logistic regression, defined as ℓ(X,Y; β) = −Y(β0 + β ⊤ 1 X) + log{1 + exp(β0 + β ⊤ 1 X)}, mb0(Xi) is as described in equation (3.10) and for a p ∈ [0, 1] the ℓ(X, p; β) = pℓ(X, 1; β) + (… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Different estimates in well-specified exponential tilt model in [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]
Figure 5
Figure 5. Figure 5: Different estimates in misspecified exponential tilt model in [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 45 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    Bang and J

    H. Bang and J. M. Robins. Doubly robust estimation in missing data and causal inference models. Biometrics, 61 0 (4): 0 962--973, 2005

    2005

  3. [3]

    Bickel, M

    S. Bickel, M. Br \"u ckner, and T. Scheffer. Discriminative learning under covariate shift. Journal of Machine Learning Research, 10 0 (9), 2009

    work page 2009

  4. [4]

    T. T. Cai and H. Wei. Transfer learning for nonparametric classification: Minimax rate and adaptive classifier. The Annals of Statistics, 49 0 (1): 0 100--128, 2021

    work page 2021

  5. [5]

    P. E. Cheng. Nonparametric estimation of mean functionals with data missing at random. Journal of the American statistical association, 89 0 (425): 0 81--87, 1994

    work page 1994

  6. [6]

    M. Das, W. K. Newey, and F. Vella. Nonparametric estimation of sample selection models. The Review of Economic Studies, 70 0 (1): 0 33--58, 2003

    work page 2003

  7. [7]

    A. P. Dempster, N. M. Laird, and D. B. Rubin. Maximum likelihood from incomplete data via the em algorithm. Journal of the royal statistical society: series B (methodological), 39 0 (1): 0 1--22, 1977

    work page 1977

  8. [8]

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248--255. Ieee, 2009

    2009

Show all 55 references
  1. [9]

    Efromovich

    S. Efromovich. Nonparametric regression with predictors missing at random. Journal of the American Statistical Association, 106 0 (493): 0 306--319, 2011

    2011

  2. [10]

    R. E. Fay. Causal models for patterns of nonresponse. Journal of the American Statistical Association, 81 0 (394): 0 354--365, 1986

    1986

  3. [11]

    S. Garg, Y. Wu, S. Balakrishnan, and Z. Lipton. A unified view of label shift estimation. Advances in Neural Information Processing Systems, 33: 0 3290--3300, 2020

    2020

  4. [12]

    K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770--778, 2016

    2016

  5. [13]

    J. J. Heckman. Sample selection bias as a specification error. Econometrica: Journal of the econometric society, pages 153--161, 1979

    1979

  6. [14]

    D. G. Horvitz and D. J. Thompson. A generalization of sampling without replacement from a finite universe. Journal of the American statistical Association, 47 0 (260): 0 663--685, 1952

    1952

  7. [15]

    D. Hu, T. Yu, and P. Li. Receiver operating characteristic curve analysis with non-ignorable missing disease status. arXiv preprint arXiv:2411.17402, 2024

    2024 arXiv

  8. [16]

    H. Jin, Y. Ma, and F. Jiang. Matrix completion with covariate information and informative missingness. Journal of Machine Learning Research, 23 0 (180): 0 1--62, 2022

    2022

  9. [17]

    M. G. Kenward and G. Molenberghs. Likelihood based frequentist inference when data are missing at random. Statistical Science, pages 236--247, 1998

    1998

  10. [18]

    J. K. Kim and C. L. Yu. A semiparametric estimation of mean functionals with nonignorable missing data. Journal of the American Statistical Association, 106 0 (493): 0 157--165, 2011

    2011

  11. [19]

    Kivinen and M

    J. Kivinen and M. K. Warmuth. Exponentiated gradient versus gradient descent for linear predictors. information and computation, 132 0 (1): 0 1--63, 1997

    1997

  12. [20]

    M. J. Laan and J. M. Robins. Unified methods for censored longitudinal data and causality. Springer, 2003

    2003

  13. [21]

    P. Li, J. Qin, and Y. Liu. Instability of inverse probability weighting methods and a remedy for nonignorable missing data. Biometrics, 79 0 (4): 0 3215--3226, 2023

    2023

  14. [22]

    Lipton, Y.-X

    Z. Lipton, Y.-X. Wang, and A. Smola. Detecting and correcting for label shift with black box predictors. In International conference on machine learning, pages 3122--3130. PMLR, 2018

    2018

  15. [23]

    R. J. Little. Regression with missing x's: a review. Journal of the American statistical association, 87 0 (420): 0 1227--1237, 1992

    1992

  16. [24]

    R. J. Little. Pattern-mixture models for multivariate incomplete data. Journal of the American Statistical Association, 88 0 (421): 0 125--134, 1993

    1993

  17. [25]

    R. J. Little. A class of pattern-mixture models for normal incomplete data. Biometrika, 81 0 (3): 0 471--483, 1994

    1994

  18. [26]

    R. J. Little and D. B. Rubin. Statistical analysis with missing data, volume 793. John Wiley & Sons, 2019

    2019

  19. [27]

    L. Liu, W. Miao, B. Sun, J. Robins, and E. T. Tchetgen. Identification and inference for marginal average treatment effect on the treated with an instrumental variable. Statistica sinica, 30 0 (3): 0 1517, 2020

    2020

  20. [28]

    Y. Liu, P. Li, and J. Qin. Full-semiparametric-likelihood-based inference for non-ignorable missing data. Statistica Sinica, 32 0 (1): 0 271--292, 2022

    2022

  21. [29]

    W.-Q. Ma, Z. Geng, and Y.-H. Hu. Identification of graphical models for nonignorable nonresponse of binary outcomes in longitudinal studies. Journal of multivariate analysis, 87 0 (1): 0 24--45, 2003

    2003

  22. [30]

    Maity, Y

    S. Maity, Y. Sun, and M. Banerjee. Minimax optimal approaches to the label shift problem in non-parametric settings. Journal of Machine Learning Research, 23 0 (346): 0 1--45, 2022

    2022

  23. [31]

    Maity, M

    S. Maity, M. Yurochkin, M. Banerjee, and Y. Sun. Understanding new tasks through the lens of training data via exponential tilting. In The Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=DBMttEEoLbw

    2023

  24. [32]

    Maity, D

    S. Maity, D. Dutta, J. Terhorst, Y. Sun, and M. Banerjee. A linear adjustment-based approach to posterior drift in transfer learning. Biometrika, 111 0 (1): 0 31--50, 2024

    2024

  25. [33]

    X. Mao, S. X. Chen, and R. K. Wong. Matrix completion with covariate information. Journal of the American Statistical Association, 114 0 (525): 0 198--210, 2019

    2019

  26. [34]

    Miao and E

    W. Miao and E. J. Tchetgen Tchetgen. On varieties of doubly robust estimators under missingness not at random with a shadow variable. Biometrika, 103 0 (2): 0 475--482, 2016

    2016

  27. [35]

    W. Miao, P. Ding, and Z. Geng. Identifiability of normal and normal mixture models with nonignorable missing data. Journal of the American Statistical Association, 111 0 (516): 0 1673--1683, 2016

    2016

  28. [36]

    W. Miao, L. Liu, Y. Li, E. J. Tchetgen Tchetgen, and Z. Geng. Identification and semiparametric efficiency theory of nonignorable missing data with a shadow variable. ACM/JMS Journal of Data Science, 1 0 (2): 0 1--23, 2024

    2024

  29. [37]

    N. G. Nair, P. Satpathy, J. Christopher, et al. Covariate shift: A review and analysis on classifiers. In 2019 Global Conference for Advancement in Technology (GCAT), pages 1--6. IEEE, 2019

    2019

  30. [38]

    J. M. Robins, A. Rotnitzky, and D. O. Scharfstein. Sensitivity analysis for selection bias and unmeasured confounding in missing data and causal inference models. In Statistical models in epidemiology, the environment, and clinical trials, pages 1--94. Springer, 2000

    2000

  31. [39]

    Rotnitzky, J

    A. Rotnitzky, J. M. Robins, and D. O. Scharfstein. Semiparametric regression for repeated outcomes with nonignorable nonresponse. Journal of the american statistical association, 93 0 (444): 0 1321--1339, 1998

    1998

  32. [40]

    D. B. Rubin. Inference and missing data. Biometrika, 63 0 (3): 0 581--592, 1976

    1976

  33. [41]

    D. B. Rubin. Multiple imputation for nonresponse in surveys, volume 81. John Wiley & Sons, 2004

    2004

  34. [42]

    Sagawa, P

    S. Sagawa, P. W. Koh, T. B. Hashimoto, and P. Liang. Distributionally robust neural networks. In International Conference on Learning Representations, 2020. URL https://openreview.net/forum?id=ryxGuJrFvS

    2020

  35. [43]

    Schenker and A

    N. Schenker and A. H. Welsh. Asymptotic results for multiple imputation. The Annals of Statistics, 16 0 (4): 0 1550--1566, 1988

    1988

  36. [44]

    C. Scott. A generalized neyman-pearson criterion for optimal domain adaptation. In Algorithmic Learning Theory, pages 738--761. PMLR, 2019

    2019

  37. [45]

    Sugiyama, M

    M. Sugiyama, M. Krauledat, and K.-R. M \"u ller. Covariate shift adaptation by importance weighted cross validation. Journal of Machine Learning Research, 8 0 (5), 2007

    2007

  38. [46]

    B. Sun, L. Liu, W. Miao, K. Wirth, J. Robins, and E. J. T. Tchetgen. Semiparametric estimation with data missing not at random using an instrumental variable. Statistica Sinica, 28 0 (4): 0 1965, 2018

    1965

  39. [47]

    Tang and Y

    N. Tang and Y. Ju. Statistical inference for nonignorable missing-data problems: a selective review. Statistical Theory and Related Fields, 2 0 (2): 0 105--133, 2018

    2018

  40. [48]

    E. J. Tchetgen Tchetgen and K. E. Wirth. A general instrumental variable framework for regression analysis with outcome missing not at random. Biometrics, 73 0 (4): 0 1123--1131, 2017

    2017

  41. [49]

    A. A. Tsiatis. Semiparametric theory and missing data, volume 4. Springer, 2006

    2006

  42. [50]

    C. Wah, S. Branson, P. Welinder, P. Perona, and S. Belongie. The caltech-ucsd birds-200-2011 dataset. 2011

    2011

  43. [51]

    L. Wang, J. Shao, and F. Fang. Propensity model selection with nonignorable nonresponse and instrument variable. Statistica Sinica, 31 0 (2): 0 647--672, 2021

    2021

  44. [52]

    S. Wang, J. Shao, and J. K. Kim. An instrumental variable approach for identification and estimation with nonignorable nonresponse. Statistica Sinica, pages 1097--1116, 2014

    2014

  45. [53]

    W. Yu, J. K. Kim, and T. Park. Estimation of area under the roc curve under nonignorable verification bias. Statistica Sinica, 28 0 (4): 0 2149, 2018

    2018

  46. [54]

    B. Zhou, A. Lapedriza, A. Khosla, A. Oliva, and A. Torralba. Places: A 10 million image database for scene recognition. IEEE transactions on pattern analysis and machine intelligence, 40 0 (6): 0 1452--1464, 2017

    2017

  47. [55]

    Z. Zhu, T. Wang, and R. J. Samworth. High-dimensional principal component analysis with heterogeneous missingness. Journal of the Royal Statistical Society Series B: Statistical Methodology, 84 0 (5): 0 2000--2031, 2022

    2000

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.